search

ReliaSolve / Molprobity2

0 stars 0 forks source link

Different numbers of suites for 3cma in -report format #157

Closed russell-taylor closed 3 years ago

russell-taylor commented 3 years ago 
taylorr@reliasolve-3:~/src/suitename_regression/outputs$ diff 3cma.report.orig 3cma.report.new 
2313c2313
<  :1: 0:2471: : :  G 33 t 1c 0.767
---
>  :1: 0:2471: : :  G 33 t 1c 0.766
2889,2890c2889,2890
<    100 suites have suiteness >=.3 <.4
<    120 suites have suiteness >=.4 <.5
---
>    101 suites have suiteness >=.3 <.4
>    119 suites have suiteness >=.4 <.5
2901,2902c2901,2902
<     34 suites have suiteness >=.3 <.4
<     46 suites have suiteness >=.4 <.5
---
>     35 suites have suiteness >=.3 <.4
>     45 suites have suiteness >=.4 <.5
russell-taylor commented 3 years ago

It sounds from @Joymaker like these are corner cases that get rounded differently, possibly due to single- vs. double-precision floating point. We'll ignore them for now.